s to come
> up with a data model that avoids the strangulation. CQL is a nice
> syntactic layer, but, at the end of the day, to avoid performance
> black holes, you have to understand how the data is going to be
> stored by Cassandra in rows and columns.
> >
>
e day,
> to avoid performance black holes, you have to understand how the data is
> going to be stored by Cassandra in rows and columns.
> >
> >
> > -Original Message-
> > From: Simon Chemouil [mailto:schemo...@gmail.com]
> > Sent: Monday, June 02, 2014 10:
understand how the data is going to be
> stored by Cassandra in rows and columns.
>
>
> -Original Message-
> From: Simon Chemouil [mailto:schemo...@gmail.com]
> Sent: Monday, June 02, 2014 10:56 AM
> To: user@cassandra.apache.org
> Subject: Re: Performance migrati
lto:schemo...@gmail.com]
Sent: Monday, June 02, 2014 10:56 AM
To: user@cassandra.apache.org
Subject: Re: Performance migrating from MySQL to C*
Hi Moshe,
Thanks for your answer and the link on time series.
We'd like to query on more than one dataName, but also on the time range and on
an arbi
Hi Moshe,
Thanks for your answer and the link on time series.
We'd like to query on more than one dataName, but also on the time range
and on an arbitrary number of sensorIds. Which we can't seem to do with
CQL because we can't have multiple IN clauses or IN clauses on the
primary key. (hopefully
Hello DuyHai,
Thanks for your answer. Unfortunately having one 'metricType' field or
storing each metric in different partitions is roughly the same for us
if we have to "merge" them on the timestamp client-side. Your solution
can reduce the number of queries which is practical but we might get
qu
Just looking at the data modeling issue:
Your queries seem to always be for a single dataName. So, that should be the
main part of the row key.
Within that, it seems you need to be able to select a range based on time. So,
time should be the primary sort key for the column name.
Based on those
Hello Simon
There is definitely some data modeling issue there.
Your first data model was quite good, except the usage of
map. Collections in C* are not meant to be used to store lots
of values because they are loaded entirely in memory server side every time
you access them. An alternative for
